Home Services Pay-Per-Call Playbook: HVAC, Plumbing, and Roofing Lead Buying in 2026

A roofer in Tampa pays $85 for a call. Two minutes in, the prospect's shopping three quotes for a job he hasn't decided to do. That's the median in this vertical — and why most home services pay per call campaigns quietly bleed margin every month.

We see this every week running VeloCalls. Home services is the most-pitched, least-understood corner of pay-per-call. Payouts look great on a rate card. Reality: HVAC, plumbing, and roofing each have their own economics, fraud profile, and TCPA traps — treating them like one bucket is the fastest way to torch a budget.

This is the playbook we'd hand an agency starting fresh in 2026. Real numbers, named publishers, opinions.

Quick admission. Last summer we told a roofer to skip Aragon for plumbing leads — good advice, except he was buying for a sister plumbing company too. We didn't ask, he didn't volunteer, and he ended up trusting a competing platform on that side. Lost the deal. Ask about adjacent verticals. Always.

Vertical Economics: What You're Actually Paying

The CPL ranges most networks won't print on their pitch deck.

HVAC — CPL: $35-65. Payout: $40-75. Buyer's effective cost after qualification: $55-90 per booked appointment. Margin per closed job (national average ticket ~$650) works if close rate clears 18%. Seasonality is brutal — July spend in Phoenix or Houston runs ~2.5x April spend, and most buyers don't pace for it.

Plumbing (emergency) — CPL: $80-150. Payouts hit $120-180 for after-hours calls in Tier 1 metros. This is where the money is. Also where the fraud is. Buyer effective cost per appointment: $110-175. Average emergency ticket clears $850, so the math works if — big if — your routing pipes calls to a tech who can be on-site in 60 minutes.

Plumbing (non-emergency) — CPL: $30-55. Payouts $35-60. Longer sales cycle, higher tire-kicker rate. Most buyers we audit shouldn't be in this bucket at all — close economics don't compete with emergency.

Roofing — CPL: $40-90. Payouts vary wildly: $50 for general inquiries, up to $120 for storm-damage and insurance-claim calls. The catch: roofing has the longest sales cycle of the three (30-60 days from call to signed contract), so attribution gets messy and buyers under-pay publishers who don't track properly.

The number that matters across all three: effective margin per call, calculated as (payout × qualification rate) − cost. If that's under $8, kill the campaign. We don't care what the rate card says.

The Buyer Landscape

Most home services pay per call traffic flows through a small set of networks. Here's what we hear from operators actually running spend.

• Service Direct — Largest US home services network. Lead quality acceptable, pricing high, support slow. Good for buyers who need volume immediately and don't mind paying tax for it.
• Aragon Advertising — Strong on insurance and finance. Skip Aragon for plumbing — quality has tanked since the 2024 acquisition, and support won't credit obviously-bad calls anymore. We've raised this with them twice. No movement.
• Inquirly — Newer, aggressive on payouts to publishers, which means better volume but rougher leads. Pilot-only until you've vetted 200+ calls. The first 50 we ran were 40% out-of-area.
• Ringba marketplace — Less a network than a buyer/seller exchange. If you have your own publisher relationships, this is where you wire them up — not where you discover them.
• Direct affiliate relationships — The actual best move once you know the vertical. Skip the 20-30% network take, sign directly with 2-3 publishers who own SEO for "emergency plumber [city]" terms.

A take we'll defend: skip Invoca unless you're past $50K/mo spend — the conversation intel is genuinely good, but the pricing isn't worth it under that volume. TrackDrive is the better mid-market choice for most buyers we audit, even though it's less polished. Under $10K/mo? Use CallRail and sample calls manually. Fancy tools are a tax until you have volume to justify them.

TCPA Gotchas That Will Eat You Alive

Home services is the second-most-litigated TCPA vertical after insurance. The rules changed materially in 2024-2025, and a lot of buyers haven't updated their compliance posture. Most don't even know they need to.

One-to-one consent. The FCC's 2024 update (survived court challenges, enforced in 2026) requires consent specific to a single seller — not a blanket "we and our partners may contact you" disclosure on a comparison site. If you're buying calls from a publisher running a "Get HVAC quotes" form, that consent is probably not legally yours. Get the record, with timestamp and IP, before you pay the invoice. We've audited buyers paying six figures a month against consent records they had never asked to see.

Mobile dialing. Express written consent is required before autodialing or sending pre-recorded messages to a mobile number. Doesn't matter that the consumer "called you first" — once you autodial them back, TCPA applies. Audit any auto-callback feature in your CRM this week.

State overlays. Florida and Washington have mini-TCPA statutes with stricter consent rules and statutory damages. National campaigns need state-aware consent capture, not a one-size-fits-all form.

Recording disclosure. Two-party consent states require explicit recording disclosure at call start. Most publisher whisper messages don't qualify. We've seen $40K judgments off a single recorded call in California.

Honest summary. TCPA enforcement in 2026 has gone from "occasional" to weekly news in the trade press. Agencies that get caught are usually the ones treating consent as a checkbox. The ones treating it as an ops requirement will quietly take their market share.

Call Routing Setup That Actually Works

Routing in home services has three dimensions, and you need all of them.

Geography. Caller's area code and IP geolocation should both fall inside the buyer's service area, with overlap rules near territory edges. About 8-12% of calls in our data come from outside the buyer's actual service area — routing them anyway burns money and erodes the relationship.

Time of day. Plumbing emergency at 2am goes to a buyer with after-hours coverage, period. HVAC service calls at 11pm hit voicemail — don't buy them. Match buyer-by-buyer availability, not a generic "business hours" filter. Holiday calendars matter too. We watched a campaign lose 6% of monthly spend routed to closed offices on July 4th.

Agent skill / job type. A roofer doing storm-damage insurance work shouldn't get routine inspection calls. A plumber who specializes in commercial shouldn't get residential drain clogs. Tag every buyer by job type. The biggest conversion lift we see in audits — usually 15-25 points — comes from skill-based routing. Not fancy AI scoring.

Layer on conversion-weighted routing once you have 30+ days of close data. Buyer closing at 28% gets 3x the volume of the one closing at 11%. Most platforms support this. Almost nobody turns it on.

For the click side of the same problem, our sister product ClickzProtect handles paid search fraud detection — pay-per-call fraud and click fraud are the same economy with different surfaces. For the deeper call-side setup, see pay-per-call campaigns bleeding money.

Top 5 Publishers Worth Working With (With Caveats)

We are not partnered with any of these. Information is from operator interviews and our own platform's call-source data.

1. Service Direct — High volume, premium pricing. Use for fast ramp; squeeze pricing once you have 60 days of data.
2. MediaAlpha (home services arm) — Strong on roofing and HVAC, weaker on plumbing. Cleaner consent records than most.
3. Networx — Home improvement broadly. Decent for non-emergency plumbing and roofing; skip for emergency. Volume's not there.
4. HomeAdvisor / Angi pay-per-call — Volume is there. Quality is inconsistent and getting worse — they're optimizing for their marketplace, not yours. We've watched buyers' close rates drop 4-6 points after Angi shifted attribution rules in 2025. Filler, not foundation.
5. Direct SEO publishers — The 2-3 affiliates who actually rank for "emergency plumber [your city]." Gold. Search the terms yourself, contact the site owner, offer a direct PPCall deal at network-minus-25%. First time we tried this for a Phoenix plumber, we got two yeses out of seven outreaches — one of those publishers is now their best source.

Skip Aragon for plumbing in 2026. Skip any network that won't show you per-publisher source breakdown. Skip any deal that bundles consent liability onto you without consent records.

Common Mistakes That Cost Real Money

Paying for short calls. Set qualification at 90 seconds minimum, 120 for plumbing emergency. Anything shorter is almost always a price-shopper or fraud signal. Buyers paying for sub-30-second calls is normal in this industry. That's just lighting money.

Not filtering by intent. A 4-minute call where the prospect says "just getting quotes for next month" is not the same as a 4-minute booked appointment. Use conversation intelligence, or sample 20% of calls manually. Filter on intent, not duration.

No fraud detection. Home services has lower fraud rates than insurance, but it's not zero — repeat callers, VOIP spoofing, duration-stuffing all show up. Without multi-signal fraud detection, you're losing 8-15% of spend you'll never see. We've onboarded buyers who didn't believe the number until we played them the recordings.

Trusting network QA. Networks have a structural conflict of interest in their own QA scoring. Sample your own calls. Always. We made this mistake ourselves with a network in 2024 — trusted their "high quality" tag the first 30 days because volume looked great, only later realized 22% of those tagged calls were sub-60-second. Expensive month.

Ignoring seasonality. HVAC in July is not HVAC in April. Flat-pace your budget and you'll overpay in shoulder seasons and run out of money exactly when calls convert best.

Where to Start

If you're standing up a home services pay per call program from zero, here's the order.

1. Pick one vertical. Plumbing emergency is the highest-margin entry point if you can support 24/7 routing. HVAC is the largest market. Roofing has the longest sales cycle and is hardest to attribute.
2. Sign with one network and one direct publisher. Run 100 calls each. Compare close rates, not duration.
3. Set up routing in three layers: geo, time-of-day, skill. Turn on capacity-based fallback.
4. Stand up consent-record collection before you pay the first invoice. Highest-leverage compliance investment you can make.
5. Sample 20% of calls manually for the first 30 days. Trust nothing the network's QA tells you until your own ears agree. For more operator-level notes, see our blog.

If you want the call tracking, fraud detection, and routing stack built for this, that's what VeloCalls does. If you'd rather build it on Ringba or CallRail, you can — most of this playbook applies regardless of platform. Pick the tools that fit your stack and run the math honestly.

Frequently Asked Questions

What are typical commission structures for home services pay-per-call?

Most networks pay flat per qualified call — qualification usually means a 90-120 second duration, in-state caller, in business hours. HVAC payouts run $35-65, plumbing $50-150 (emergency leans high), roofing $40-90. Some networks offer revshare on closed jobs (5-15% of ticket), but those deals are rare and usually reserved for top publishers. Avoid hybrid CPA/PPCall structures unless you trust the buyer's reporting — they almost always under-attribute.

How do I verify lead quality before scaling spend?

Run a 100-call pilot before committing. Track three things per call: caller intent (real job vs. price-shopping), in-service-area rate, and 30-day close rate from the buyer. If close rate is below 12% for HVAC or 18% for plumbing emergency, the lead source is bad — doesn't matter what the call duration says. Pull 10 random call recordings yourself; do not trust the network's QA scoring.

What's my TCPA exposure as a pay-per-call buyer?

Real, and growing. Under the FCC's 2024 one-to-one consent rule (effective and now being enforced in 2026), each buyer needs documented express written consent specific to them — not a blanket consent on a comparison site. If you're buying calls through an affiliate that runs a lead-gen form, get the consent records on file before paying a single invoice. Settlements in home services TCPA class actions have hit $5-12M; this is not a theoretical risk.

How long does it take to ramp a new home services vertical?

Plan on 60-90 days to know if a vertical works for you. Weeks 1-2 are buyer setup and routing config. Weeks 3-6 you're running pilot volume and tuning fraud filters. Weeks 7-12 you scale the publishers that hit your margin target and cut the rest. If anyone tells you they ramped a new vertical to profitability in 30 days, they either had pre-existing buyer relationships or they're not counting properly.

What's the best CRM setup for pay-per-call?

For home services specifically: Service Titan or Housecall Pro on the buyer side, plus a call tracking layer (VeloCalls, Ringba, or CallRail) that pushes call metadata into the CRM via webhook. You want every call logged with source, duration, recording URL, and disposition — not just calendar appointments. The agencies that win this are the ones who close the loop from call event to closed ticket inside 24 hours.